ORIGINAL ARTICLE

Exploring functional contexts of symbiotic sustainwithin lichen-associated bacteria by comparativeomics

Martin Grube1,5, Tomislav Cernava2,5, Jung Soh3,5, Stephan Fuchs4, Ines Aschenbrenner1,2,Christian Lassek4, Uwe Wegner4, Dorte Becher4, Katharina Riedel4, Christoph W Sensen3

and Gabriele Berg2

1Institute of Plant Sciences, Karl-Franzens-University, Graz, Austria; 2Institute of EnvironmentalBiotechnology, Graz University of Technology, Graz, Austria; 3Department of Biochemistry & MolecularBiology, University of Calgary, Calgary, Alberta, Canada and 4Institute of Microbiology, Ernst-Moritz-ArndtUniversity of Greifswald, Greifswald, Germany

Symbioses represent a frequent and successful lifestyle on earth and lichens are one of their classicexamples. Recently, bacterial communities were identified as stable, specific and structurallyintegrated partners of the lichen symbiosis, but their role has remained largely elusive incomparison to the well-known functions of the fungal and algal partners. We have explored themetabolic potentials of the microbiome using the lung lichen Lobaria pulmonaria as the model.Metagenomic and proteomic data were comparatively assessed and visualized by Voronoi treemaps.The study was complemented with molecular, microscopic and physiological assays. We havefound that more than 800 bacterial species have the ability to contribute multiple aspects to thesymbiotic system, including essential functions such as (i) nutrient supply, especially nitrogen,phosphorous and sulfur, (ii) resistance against biotic stress factors (that is, pathogen defense),(iii) resistance against abiotic factors, (iv) support of photosynthesis by provision of vitamin B12,(v) fungal and algal growth support by provision of hormones, (vi) detoxification of metabolites, and(vii) degradation of older parts of the lichen thallus. Our findings showed the potential of lichen-associated bacteria to interact with the fungal as well as algal partner to support health, growth andfitness of their hosts. We developed a model of the symbiosis depicting the functional multi-playernetwork of the participants, and argue that the strategy of functional diversification in lichenssupports the longevity and persistence of lichens under extreme and changing ecologicalconditions.The ISME Journal (2015) 9, 412–424; doi:10.1038/ismej.2014.138; published online 29 July 2014

Introduction

Symbiosis, one of the most common lifestyles onearth, is a long-term interaction, which acts as sourceof evolutionary innovation (Margulis and Fester,1991; Martin and Schwab, 2013). The term symbiosiswas introduced by Frank (1877) in a study of lichens,which are today considered a classic examplefor self-sustaining partnerships of species belongingto different kingdoms of life (Nash, 2008).Lichens represent one of most diversified and oldestsymbiotic lifestyles of fungi, with more than 18 000recognized fungal species and a typical stratified

morphology that evolved at least 415 million yearsago (Honegger et al., 2013). The light-exposed lichenthalli are shaped by outer layers of fungal hyphae,which shelter internalized phototrophic partners.Lichen thalli develop only with proper combinationsof fungal and algal species. Once the symbioticphenotype is established, lichens may reach indeter-minate ages and may even survive even the harshestconditions on earth (Øvstedal and Lewis-Smith,2001). One reason for the ecological success of thisfungal-algal partnership is the mutually enhancedability to survive oxidative stress by suspendedanimation and rapid resumption of metabolismunder permissive conditions (Kranner et al., 2005).However, additional and hitherto unidentified forcesmight have helped lichens to adapt to nutrient-poorand hostile habitats with strong fluctuation of abioticparameters.

Most biology textbooks characterize lichens as anassociation solely between a fungal (mycobiont) and

Correspondence: G Berg, Institute of Environmental Bio-technology, Graz University of Technology, Petersgasse 12,8010 Graz, Austria.E-mail: gabriele.berg@tugraz5These authors contributed equally to this work.Received 7 March 2014; revised 13 June 2014; accepted 17 June2014; published online 29 July 2014

The ISME Journal (2015) 9, 412–424& 2015 International Society for Microbial Ecology All rights reserved 1751-7362/15

www.nature.com/ismej

an algal (photobiont) partner. Our recent studies,however, revealed a high diversity of bacteriathat are also associated with lichens. We havevisualized biofilm-like communities, dominatedby Alphaproteobacteria on the surfaces of thalli(Cardinale et al., 2008, 2012a, b). Their structuresuggests some degree of host specificity of thebacterial communities (Grube et al., 2009;Hodkinson et al., 2011; Bates et al., 2011). Theubiquity and abundance of lichen-associated bacteriachallenge the classic view of lichens as a two-tierpartnership and support an extended concept thatalso includes the universally present bacterial com-munities. With this new perspective in mind, wehave now explored the potential functions of thebacterial communities with the goal of understandingtheir symbiotic context in a more holistic way.Functional analysis of the culturable fraction of thelichen-associated bacteria suggested their involve-ment in several roles, including iron and phosphatemobilization, hormone production, nitrogen fixationas well as several lytic activities (Liba et al., 2006;Grube et al., 2009). The culturable fraction usuallyrepresents only a minor part of the total lichen-associated microbiome and therefore most likely doesnot cover the host-dependent majority of the bacterialspecies (Cardinale et al., 2008). Recently developedomics approaches and subsequent bioinformaticstools are therefore required for a more comprehensiveunderstanding of the role of the microbiome.

The objective of our study was therefore toinvestigate the function and metabolic potential ofthe bacterial lichen microbiome using a combinedomics approach, together with a comprehensivespectrum of molecular, microscopic and physiolo-gical assays. The results were comparativelyassessed and visualized using bioinformatics tools.For our study, we chose the lung lichen Lobariapulmonaria (L.) Hoffm., a lichen which is consid-ered to be endangered in many areas. L. pulmonariaserves as an indicator of primeval forest ecologicalcontinuity (Scheidegger and Werth, 2009). Themycobiont of L. pulmonaria engulfs a green-algalphotobiont (Dictyochloropsis reticulata; found in90% of the lichen surface) and a minor cyanobacter-ial partner, Nostoc sp. (Cornejo and Scheidegger,2013). Schneider et al. (2011) provided an initialinsight into the L. pulmonaria proteome andsuggested Lobaria as an ideal model to studysymbiotic processes. In our study, we extended thiswork and now provide results from the comparisonof metagenomic and metaproteomic data.

Material and methods

Sampling strategy and preparationL. pulmonaria was sampled from a rich populationon maple tress (Acer spp.) in the Alps (Johnsbach,Austria; N 4713203500, E 1413703800) after visualinspection to avoid contamination by lichenicolous

fungi and other organisms (Supplementary FigureS1). Using integrated sampling, a total amount of176.3 g lichen was shock frozen with liquid nitrogenand immediately ground with mortar and pestle.The sample was homogenized in 360 ml 0.85% NaCland filtered using a 63 mm mesh sieve; larger lichenparts were retained and colonizing bacteria wereenriched in the filtrate. The filtrate was centrifugedat 8,000 r.p.m. at 4 1C for 20 min and the pelletwas resuspended in 16� 1.5 ml 0.85% NaClaliquots. After a subsequent centrifugation step at13 000 r.p.m. at 4 1C for 20 min, the supernatant wasdiscarded and the pellets were used for DNAisolation (PowerSoil DNA Isolation Kit, MO BIOLaboratories Inc., Carlsbad, CA, USA). Followingthe DNA isolation, an aliquot containing 22 mg ofmetagenomic DNA was sent to GATC Biotech(Konstanz, Germany) for Illumina sequencing(HiSeq 2000 paired-end runs, Illumina Inc., SanDiego, CA, USA).

Quality control and assembly of Illumina readsIllumina HiSeq 2000 paired-end metagenomic DNAsequencing reads (GATC Biotech) were initiallyquality-checked using the FastQC program. Theadapter sequence (50-GATC GGAA GAGC ACACGTCT GAAC TCCAG TCAC GTCC GCAC ATCTCGTAT-30: identified as part of the Illumina TruSeqIndexed Adapter) was found in over 1% of the readsin set 1, and diminishing quality scores wereobserved towards the end of reads. Based on thisobservation, quality trimming and filtering wasperformed on each raw read set, using a custom-developed Perl script (written in-house by JS). Theadapter sequences were removed and the sequenceends were trimmed, when the base quality score wassmaller than 20. After the trimming step, reads werefiltered out (i) if the length was shorter than 75 bp;(ii) or the read contained one or more ambiguous (N)bases; or (iii) the average quality score overall bases ofthe read was o25. After quality control, we did notobserve overrepresented sequences, significantdegradation of base qualities, or any other majorquality issues. The sequence quality was confirmedagain using the FastQC program. After qualitycontrol, some reads from set 1 did not have matchingreads from set 2 due to read filtering, and vice versa.Therefore, to use paired-end information duringassembly, only those reads that formed a matchingpair were retained for further analysis.

Assembly of the reads into contiguous sequences(contigs) was performed using the Velvet de novoassembly software (Zerbino and Velvet, 2008;http://www.ebi.ac.uk/Bzerbino/velvet/). Multipletrial assemblies with different k-mer lengths (57, 61,65, 67, 69, 71 and 73) were performed and assemblystatistics were compared with find the best k-merlength. All assemblies were conducted with an insertlength of 350. The final k-mer length chosen was 71,which resulted in the maximum N50 value (2411 nts).

Symbiotic sustain within lichen-associated bacteriaM Grube et al

413

The ISME Journal

Taxonomic and functional analysis of assembledIllumina readsThe number of actual contigs used for this analysiswas 368 424 (out of 503 528 overall contigs), basedon the default Velvet minimum contig-length thresh-old of two times the k-mer length and the defaultcoverage cutoff of half the median coverage.The Tera-BLASTN program (www.timelogic.com/documents/TeraBLAST_2009.pdf) was run on the368 424 contigs, using TimeLogic (Active Motif,Carlsbad, CA, USA) DeCypher boards against the‘nt’ database from NCBI (ftp://ftp.ncbi.nlm.nih.gov/blast/db). The blastn results were imported intoMEGAN (Metagenome Analyzer, v4.70.4; Husonet al., 2011) to produce several taxonomy profiles.For assembly-based functional analysis, we used asimilar approach as above but using BLASTX(www.timelogic.com/documents/TeraBLAST_2009.pdf), which was run against the ‘non-redundantprotein sequence’ database from NCBI (ftp://ftp.ncbi.nlm.nih.gov/blast/db). The blastx results wereimported into MEGAN (v4.70.4) for functional analysis.Both SEED and KEGG functional analyses wereconducted with MEGAN.

MG-RAST analysis of Illumina readsUnprocessed Illumina HiSeq 2000 reads wereuploaded on the MG-RAST v3 public server(Meyer et al., 2008) to undergo paired-end readsjoining and quality filtering (with default settings).67 731 962 (88.8%) out of 76 310 051 sequencespassed quality control; therein 60 015 088 (78,7%)sequences contained predicted proteins of eitherknown or unknown function, while 1 788 100 (2.3%)sequences contained ribosomal RNA genes. SEEDSubsystems Annotation was conducted with amaximum e-value cutoff of 1e�5 and a minimum60% identity cutoff. Rarefaction analysis based onidentified ribosomal RNA genes was done usingbest-hit classification and the Greengenes database(http://greengenes.lbl.gov) as the annotationsource (with a minimum e-value cutoff of 1e� 5;Supplementary Figure S2). The Lobaria metagen-ome was compared within a Principal CoordinatesAnalysis (annotation source: subsystems) with 20publicly available datasets on MG-RAST. The func-tional abundance of eight particular habitats wascompared using a minimum e-value cutoff of 1e� 5.A table with all compared habitats together withtheir MG-RAST accession numbers is provided inthe Supplementary Material (Supplementary TableS1). The metagenomic dataset is available underMG-RAST ID 4530091.3.

Quantitative real-time PCRQuantification of nifH genes in the lichen DNAextract was conducted with primer pair nifH-F/nifH-R, as described by Hai et al. (2009). Standardscontaining the nifH fragments were prepared

according to Bragina et al. (2013). Briefly, the genefragments from Erwinia carotovora subsp. atro-septica SCRI1043 were cloned into the pGEM-TEasy Vector (Promega, Madison, WI, USA) andlater re-amplified with vector-specific primers.Amplification-grade DNase I (Sigma-Aldrich, St Louis,MI, USA) treated total DNA extract was usedto determine inhibitory effects of co-extractedsubstances. Based on this experiment, the totalcommunity DNA was diluted to 1:25 and targetgenes were amplified using KAPA SYBR FASTqPCR Kit (Kapa Biosystems, Woburn, MA, USA).Two independent runs, with three replicates foreach sample, were performed on the Rotor Gene6000 (Corbett Research, Mortlake, VIC, Australia),according to Bragina et al. (2013). The specificity ofthe amplicons was confirmed with both melting-curve analysis and gel electrophoresis of the qPCRproducts, respectively. Gene copy numbers for nifHwere calculated per gram of lichen fresh weight.

Sample preparation for protein extractionThe collected thalli (Johnsbach, Austria; N 4713203500,E 1413703800) were cleaned with sterile tweezersfrom moss, bark and other visible contaminations.Samples from different thalli were pooled toa total amount of 2 g. Liquid nitrogen was addedto the pooled thalli, which were subsequentlyground to a fine powder using mortar andpestle. Proteins were extracted as described by Wanget al. (2006).

Gel electrophoresis with extracted proteinsOne-dimensional SDS-polyacrylamide gel electro-phoresis was performed as described earlier(Laemmli, 1970), and by loading 50 mg of extractedlichen protein mixture per lane. The sample wasanalyzed in three technical replicates (three lanes).Electrophoresis was carried out at 150 V and 250 mAfor 45 min, afterwards proteins were fixed byshaking the gel in an aqueous solution containing40% ethanol and 10% acetic acid for 30 min.Finally, proteins were stained with 25 ml of colloi-dal Coomassie Brilliant Blue G (Sigma-Aldrich,Steinheim, Germany), as described earlier (Neuhoffet al., 1988). The gel was scanned on a standard lightscanner (Microtek, Hsinchu, Taiwan) for documen-tation. Afterwards each of the three lanes was cutinto 20 pieces, as shown in the SupplementaryMaterial (Supplementary Figure S3).

In-gel digestion after SDS-polyacrylamide gelelectrophoresisEach of the 20 gel pieces from the three technicalreplicates was cut into small cubes (1 mm3), whichwere destained by adding 700 ml of 30% acetonitrilecontaining 0.2 M NH4HCO3, and shaking for 15 minat 37 1C and 1500 r.p.m. This step was repeated

Symbiotic sustain within lichen-associated bacteriaM Grube et al

414

The ISME Journal

twice. After destaining, the pieces were dried in avacuum concentrator (Eppendorf, Hamburg, Germany).In-gel digestion was performed with 50-ml-modifiedsequence-grade trypsin (Promega), with an enzymeconcentration of 2 mg ml�1 overnight at 37 1C.Subsequently, peptides were eluted by sonicationfor 15 min (Ultra sonic cleaner, VWR, Darmstadt,Germany). The eluted peptides were transferred intovials, dried in a vacuum concentrator (Eppendorf)again, and finally dissolved in 10 ml of ultrapurewater.

Mass spectrometry analysisPeptide mixtures resulting from in-gel tryptic clea-vage were subjected to liquid chromatography-tandem mass spectroscopy measurements, usingan EASYnLC 1000 (Thermo Scientific, Odense,Denmark), with self-packed columns (Luna 3m C18(2)100 A, Phenomenex, Aschaffenburg, Germany) in aone-column setup online, coupled to an OrbitrapElite mass spectrometer (Thermo Fisher, Bremen,Germany). Samples were loaded and desalted in0.1% acetic acid, with a flow rate of 700 nl min� 1,followed by peptide separation achieved by a binarynon-linear 170 min gradient from 5–50% acetoni-trile in 0.1% acetic acid at a flow rate of300 nl min�1. Mass spectrometric measurementwas performed in the Orbitrap Elite (Thermo FisherScientific) at a spray voltage of 2.4 kV applied to theemitter. After a survey scan in the Orbitrap(R¼ 30 000) tandem mass spectroscopy data wererecorded for the 20 most intensive precursor ions inthe linear ion trap. Singly charged ions were nottaken into account for tandem mass spectroscopyanalysis. The lock mass option was enabledthroughout all analyses.

Database construction for metaproteome analysis,database search and metaproteome data analysisDue to the lack of translated metagenomic sequencesof L. pulmonaria, a synthetic metagenomic-baseddatabase was created. This database contained allprotein sequences available in the NCBI NR proteindatabase (state of 13.09.26) which have beenpredicted to be present in the sample based onsimilarity searches of the metagenomic reads byBLAST (Altschul et al., 1990). Because neitherthe fungal nor the algal symbiont genomes ofL. pulmonaria have been sequenced yet, all avail-able NCBI protein sequences from fungi and algaewere added to the database, together with a setof typical contaminations (for example, porcinetrypsin, human keratin). Entries with the sameprotein sequence, but different headers werecombined into one entry by an in-house php script(written by SF). The final database contained2 473 550 protein sequences, pointing to 2 581 850GenBank identifiers (bacteria: 1 564 300; algae:

226 723; fungi: 513 152; archaea: 26 952; otherunclassified: 250 723).

The raw files were converted to mgf-files byMSconvert (www.Proteowizard.org), and searchedwith the Mascot search engine (version 2.2.04,Matrix Science Inc., Boston, MA, USA) with thefollowing parameters: parent mass tolerance10 ppm, fragment mass tolerance 0.5 Da, maximummissed cleavages 2; charge state 1þ ; variablemodifications and oxidation of methionine. TheMascot search was followed by an X-tandemanalysis in Scaffold (version 4.0.7, ProteomeSoftware Inc., Portland, OR, USA). This analysiswas performed as a ‘MudPit experiment’ to mergethe individual mascot result files into a single file.The results were filtered as follows: 99% peptideprobability, 1 peptide, 99% protein probability.Only proteins detected in two out of three technicalreplicates were considered for further analyses.

For functional classification and taxonomicaldistribution the in-house developed metaproteomeanalyses pipeline ‘Prophane 2.0’ was used (http://www.prophane.de; Schneider et al., 2011). Briefly,peptide to protein matches were clustered in groupsby the Scaffold software. To standardize functionalannotation all peptide to protein matches werefunctionally characterized based on TIGRFAMs(release 12; Haft et al., 2013) using HMMER3(e-valuep1E� 10; Haft et al., 2013; Eddy, 2011).Functional data were transferred to the proteingroup if the members share the same predictedfunction. If these proteins share multiple predictedfunctions the function with the lowest overalle-value were assigned to the respective proteingroup. Groups of proteins sharing no functionalprediction were named heterogeneous on functionallevel.

Moreover, protein groups were taxonomicallyclassified based on the annotation of the respectiveprotein members. Protein quantification was basedon normalized spectral abundance factor values(Zybailov et al., 2006), however, only exclusivespectral counts were considered.

Voronoi treemapsVoronoi treemaps were generated using Paver(Decodon, Greifswald, Germany, http://www.decodon.com/).

Fluorescence In Situ Hybridization and ConfocalLaser Scanning Microscopy samples were collectedfrom the same Lobaria population used for themetagenomics and metaproteomics analysis. Lichenthalli were fixed with 4% paraformaldehyde/phosphate-buffered saline (PBS) (v/v, 3:1) at 4 1Cfor at least 4 h, followed by three washing steps withice-cold PBS. The samples were stored at � 20 1C inethanol absolute/PBS (v/v, 1:1). Before the hybridi-zation step of the thallus, 30-mm thick cryosectionswere made. Fluorescence In Situ Hybridization wascarried out in tubes, as outlined in Cardinale et al.,

Symbiotic sustain within lichen-associated bacteriaM Grube et al

415

The ISME Journal

(2008) and samples were analyzed using a Leica TCSSPE microscope (Leica Microsystems, Heidelberg,Germany), as well as image analysis and three-dimensional modeling, using the software ImageSurfer (http://imagesurfer.cs.unc.edu/) and Imaris7.0 (Bitplane, Zurich, Switzerland), respectively.

Isolation of lichen-associated bacteriaLichen thalli of L. pulmonaria were sampled from threedifferent locations in Austria (Tamischbachgraben,N 4713204000, E 1413703500, Johnsbach, N 4713203500,E 1413703800, and St. Oswald, N 4614405000, E 1510402600).After grinding lichen samples with mortar andpestle, a homogenate was prepared using sterile0.85% NaCl in a 1:10 (w/v) ratio, together with a labstomacher (BagMixer; Interscience, St Nom, France).Diluted fractions were plated on R2A agar (Carl Roth,Karlsruhe, Germany), R2A agar with 25mg ml�1

cycloheximide, starch casein agar (Kuester andWilliams, 1964) and ISP2 agar (Shirling andGottlieb, 1966). Bacterial colonies were randomlypicked within 5 days of incubation at roomtemperature.

Screening of isolates for in vitro antagonistic activitytowards particular bacteria and fungiDual-culture experiments were carried out as con-frontation assays, using different media and targetorganisms according to Berg et al. (2002) andOpelt et al. (2007). Lichen-associated isolates werespotted on solid media pre-inoculated with E. coliXL1 and S. aureus ATCC 25923 and assessed forinhibition zones after 4 days of incubation at 30 1C.Antagonistic activity against the fungus Botrytiscinerea Pers. (TU Graz culture collection, Graz,Austria) was tested by dual culture on Waksmanagar, according to Berg et al. (2002) and assessedafter 5–7 days of incubation at 20 1C. Cultures of thelichen-colonizing fungus Rhinocladiella sp. (TUGraz culture collection) were homogenized andresuspended in sterile 0.85% NaCl. Aliquots fromone batch (50 ml) were used to inoculate each well of24-well plates, containing solid potato dextrose agar(Carl Roth, Karlsruhe, Germany). Subsequently, 100-ml culture filtrate obtained from each lichen-asso-ciated isolate was added to particular wells. After 3weeks of incubation, the wells were checked forgrowth reduction. All experiments were conductedwith replicates and carried out twice.

Functional assays with Lobaria-associated bacteriaAltogether, 388 randomly selected bacterial cultureswere subjected to functional assays based ondifferent growth media. Protease, b-glucanase andchitinase activity were analyzed according to Berget al. (2002), and phosphate solubilization of strainsas described by Nautiyal (1999). Chromobacteriumviolaceum CV026 was used to detect C4–C6

AHL-mediated quorum sensing by visualizingpurple pigmentation of the reporter strain(McClean et al., 1997). Pseudomonas putida F117pAS-C8 and P. putida F117 pAS-C12 (Steidle et al.,2001) were used to detect C8 AHLs and C12 AHLs,respectively. Visualization of the green fluorescentprotein-based AHL sensor was achieved throughepifluorescence imaging, using an Universal HoodIII (Bio-Rad, Hercules, CA, USA). All strains wereincubated at 30 1C for 48 h.

Results

Taxonomic structure of the bacterial lichen microbiomeThe analysis of 368 424 contigs revealed the taxo-nomic profile represented in Figure 1 that shows theoverall composition of the metagenome and a moredetailed structure of the dominant bacterial taxawithin. Among the Proteobacteria, Alphaproteobac-teria was the prominent taxon, with Rhizobiales andSphingomonadales as the most frequently calledorders. Within Rhizobiales, Methylobacteriaceaeand Bradyrhizobiaceae are prominent, with Rhizo-biaceae, Beijerinckiaceae, Xanthobacteriaceae andPhylobacteriaceae in minor quantities. Almost all ofthe Sphingomonadales belonged to the Sphingomo-nadaceae, and we estimate that a total of more than800 bacterial species represent the diversity of theassociated bacterial community, according to therarefaction analysis (Supplementary Figure S2).Results derived from the present metaproteomicanalysis presented an outmost similar communitystructure to the metagenome analysis. Proteo-bacteria were the predominant phylum andaccounted for 361 distinct database hits, followedby Cyanobacteria with 47 hits and Acidobacteriawith 28 hits. A complementary Fluorescence In SituHybridization visualization with Alphaproteobacteriaand Betaproteobacteria specific probes, together withunspecific eubacteria probes illustrated and generallyconfirmed the taxonomic distribution observed by themetagenome analysis. Alphaproteobacteria werepredominant and widespread on both, the upperand lower surface of the leaf-like lichen thallus,respectively, while Betaproteobacteria were lessabundant and locally restricted (Figure 2a).

Metagenome and proteome of the bacterial lichenmicrobiomeFunctional analyses of the lichen metagenomefocused on bacterial contigs, as the samplingprocedure was designed to enrich the bacterialmetagenome, using both SEED and KEGG functionalanalyses. SEED functional analysis was used forfinding a set with functions of interest (that is,carbohydrate, virulence, cofactors and so on), manyof which are standard SEED functional terms. Out ofthe 368 424 contigs, 69 823 were assigned to afunctional term and the breakdown of the assign-ments at the top level is shown as a bar graph

Symbiotic sustain within lichen-associated bacteriaM Grube et al

416

The ISME Journal

(Supplementary Figure S4). A significant number ofcontigs represented primary metabolic functions ofbacteria (amino acids and derivatives: 6440; centralcarbohydrate metabolism: 2770; fatty acids, lipidsand isoprenoids: 2721). Taxonomic separation offunctional assignments revealed that up to 50%of different functional terms were covered byAlphaproteobacteria, while less-abundant bacteriaaccount for the remaining contigs (Figure 3).Moreover MG-RAST visualization of functionalabundance with SEED subsytems annotationshowed presence of all level 4 functional terms.Out of the 368 424 contigs, 66 739 were assigned toKEGG pathways. Although the KEGG pathway

assignment is primarily developed using mammaliangenome information, it covers central capabilities ofuniversal primary metabolism (carbohydrate meta-bolism: 12 823, energy metabolism: 7616, lipidmetabolism: 5301, nucleotide metabolism: 5909,amino acid metabolism: 12 784, metabolism of otheramino acids: 4068, glycan biosynthesis and meta-bolism: 2181 and genetic information processing:12 258). Principal Coordinates Analysis carried outwith MG-RAST revealed a unique functional dis-tribution most similar to the one found on the plantphyllosphere (Figure 4; Supplementary Table S1).

Lichen samples used for metagenome analyseswere investigated in parallel on metaproteomics

Figure 1 Taxonomic spectrum visualized with Krona (www.krona.sourceforge.net/) of contigs in the metagenome for all domains of life.Circles represent taxonomic classifications in ascending order up to the family level (outermost circle). Less-abundant taxa are listedoutside the charts together with their relative abundance.

Symbiotic sustain within lichen-associated bacteriaM Grube et al

417

The ISME Journal

level. For this a in-gel tryptic digestion followed byliquid chromatography-tandem mass spectrometrybased metaproteomics approach was used as recentlydescribed (Schneider et al., 2011) combined with adatabase containing all BLAST protein hits of themetagenomic reads (best hit per read). In total, ourmetaproteome analyses revealed 4405 differentproteins, which were covered by at least onepeptide-to-spectrum match (99% peptide probabil-ity, 99% protein probability) in at least two of threereplicates. All proteins were divided into 3226

groups, based on shared peptide-to-spectrum matchesby the Scaffold software (Koskinen et al., 2011).Out of these, 2676 and 541 groups could beunambiguously assigned to a eukaryotic and eubac-terial origin, respectively, employing the Prophaneworkflow (Schneider et al., 2011). As our studyaims for a better understanding of structureand functionality of the bacterial lichen micro-biome, further functional analyses based onTIGRFAMs (e-valuep1E� 10) focused exclusivelyon protein groups of bacterial origin. Functional

Figure 2 (a) Leaf-like thallus visualization of bacteria on a cross-section by 3D reconstruction of FISH image stacks. Eubacteria (red) andAlphaproteobacteria (yellow) were found widespread on both, the upper and the lower cortex, while Betaproteobacteria (pink) were lessabundant and locally contained. Fungal hyphae (blue) and algae located under the upper cortex (green) were visualized without specificFISH probes, due to the naturally occurring fluorescence of the internal structures. (b) Model of the lichen symbiosis depicting thefunctional network of the participants. The model includes relevant functions of the colonizing bacteria, which are derived frommetagenomic/metaproteomic analysis, as well as cultivation-dependent experiments.

Figure 3 Distribution of particular functions prior and after exclusion of less-abundant taxa. The number of contigs assigned to allorganisms within the sample (blue bars) is visualized in contrast to assigned contigs of the eight most-abundant taxa(Alphaproteobacteria, Betaproteobacteria, Gammaproteobacteria, Deltaproteobacteria, Actinobacteria, Acidobacteria, Bacteroidetesand Firmicutes; red bars) and Alphaproteobacteria (green bars). The full colour version of this figure is available at ISME Journal online.

Symbiotic sustain within lichen-associated bacteriaM Grube et al

418

The ISME Journal

data were only transferred to bacterial proteingroups whose members share the same functionalprediction (424 out of 541, see material and methodsfor details). To gain a view on the role of bacterialproteins, the data were visualized with Voronoitreemaps, which also show the participation ofdifferent eubacterial groups in functions (Figure 5).In addition, we integrated the metaproteomic datawith metagenomic data using the Voronoi treemap(Figure 6). Generally, most of the functional classespredicted by the metagenome analysis are alsorepresented by at least one protein. As expected,numerous proteins assigned to functions involvedin protein synthesis, DNA replication, recombina-tion and repair were found in the metaproteome.Moreover, a significant number of proteins involvedin protein fate and central energy metabolism, forexample, tricarbonic acid cycle, were recoveredfrom the bacterial lichen microbiome.

Selected functions of the bacterial lichen microbiome

Nutrient supply. About 2793 contigs suggested thepresence of Ton and Tol transport systems, some ofwhich are also involved in iron uptake. The TonB-dependent receptor, a family of beta-barrel proteinsfrom the outer membrane of Gram-negative bacteriaand responsible for siderophore transport into theperiplasm, was present in 2094 contigs. Meta-proteome analysis indicated at least four differenttypes of TonB-dependent receptors, which were derived

from two different bacterial phyla (Proteobacteriaand Bacteroidetes, respectively). Phosphate metabolismis represented in 885 contigs, as well as in two PFAMclassifications within the metaproteome and alsoincludes proteins involved in solubilization of phos-phates. Corresponding to this finding, 19.6% of allbacterial isolates from Lobaria formed clearing zoneson NBRIP agar and thus underscored the potentialrole of lichen-associated bacteria in the solubilizationof phosphates. Moreover, we utilized a quantitativereal-time PCR approach, based on total communityDNA, to evaluate the bacterial potential for nitrogenfixation. Therewith log10 5.0±0.1 nifH copies wereidentified per gram of lichen fresh weight.

Resistance against biotic stress factors (pathogendefense). Virulence functions are common amongthe lichen-associated bacteria. About 1152 contigswere assigned to multidrug resistance efflux pumps,and multiple genes that code for resistance againstantibiotics were also found (fluoroquinolone,vancomycin, methicillin, penicillin and cephalo-sporine) in addition to the 955 contigs that contribute

Figure 4 Principal Coordinates Analysis (PCoA) including 20publicly available metagenomic datasets from MG-RAST and theLobaria metagenome (red dot). All datasets were compared withsubsytems and calculated using normalized values and the Bray-Curtis distance matrix. Single metagenomes from different biomesare labeled with their MG-RAST accession numbers and groupedin colored ellipses. The full colour version of this figure isavailable at ISME Journal online.

Figure 5 Voronoi Treemap visualization of the prokaryoticmetaproteome on the taxonomic (a) and functional (b) level.The taxonomic and functional distributions were carried out byProphane 2.0. Taxonomic groups are indicated by different colors(a and b) and the functional classes are separated by black lines(based on TIGRRoles). Each cell represents at least one protein (orprotein group) assigned to the respective phylum.

Symbiotic sustain within lichen-associated bacteriaM Grube et al

419

The ISME Journal

to Type III, Type IV, Type VI and ESAT secretionsystems. Type VI secretion systems were representedby three different variants in the metaproteome data,as well as one specific hit for an acriflavin-resistanceprotein. A moderate number of genes were alsofound which contribute to secondary metabolism,according to SEED analysis (548 contigs); some of

the secondary metabolites are known as antibacter-ials or antifungals, but are represented by a fairlylow number of contigs (phenazine biosynthesisprotein PhzF: 26 contigs, clavulanic acid biosynthesis:11 contigs). Furthermore, results from cultivation-dependent studies involving Lobaria-associatedbacteria combined with taxonomic data from thepresent metagenomic approach suggest that around7% of the taxa present have an antagonisticpotential. The majority of these showed antagonismagainst fungi (data not shown).

Resistance against abiotic stress factors. Stress-related functions are distinct in the bacterial micro-biome (2769 contigs; oxidative stress: 1238). About891 contigs indicated genes conferring resistance tometals (copper, cobalt-zinc-cadmium, silver, mer-cury and arsenic). Interestingly, a similar proportionof oxidative-stress protectants to heavy metal effluxpumps was observed in the metaproteome (3 and 1,respectively). In addition, we looked for genesinvolved in dormancy and sporulation, as weexpected this to be an important aspect of bacteriaadapted to the poikilohydric life on lichens. How-ever, only 13 contigs were related to this function.

Photosynthesis support by vitamin B12. Numerousgenes are involved in the metabolism of cofactors,vitamins and prostethic groups (3799 contigs),according to SEED functional analysis, and arelatively high number of 1203 contigs was anno-tated to tetrapyrrole biosynthesis. Among these, forexample, 365 contigs coded for coenzyme-B12

biosynthesis, 312 contigs for thiamine and 174contigs indicated biotin biosynthesis. In the KEGGanalysis, 5479 contigs were linked with the meta-bolism of cofactors and vitamins. Metaproteomeanalysis resulted in 2 hits that support the presenceof enzymes involved in cobalamin biosynthesis,while folic acid and thiamine biosynthesis wererepresented by 1 hit each, respectively.

Hormone production. For the potential productionof hormones, we found that auxin biosynthesis wasrepresented by 156 contigs.

Detoxification of metabolites. In the KEGG analy-sis, many of the contigs that indicate xenobioticsbiodegradation and metabolism (5482 contigs) seemto also be involved in the degradation of phenoliccompounds. The presence of enzymes involved indegradation of phenolic compounds was reinforcedby three hits in the metaproteome data.

Lytic activities. Degradation of older thallus partscould provide nutrients for the young and growingthallus parts. Such function is enhanced by bacterialenzymes involved in specific degradation proce-dures. 341 contigs within the metagenome areconnected to chitin and N-acetylglucosamine utili-zation. Additionally, 1000 contigs were found for

Figure 6 Functional recovery of the bacterial lichen microbiomemetagenome on the metaproteomic level using Voronoi treemaps.The protein sequences were compared with metagenomic contigsby BLAST and then functionally characterized using HMMER3and TIGRFAMs. Upper panel shows all functions covered by themetagenome (third level) and the respective subroles (secondlevel) and main roles (first level). The mainrole labels are shown.On the lower panel, metaproteomic coverage is shown (blue cells:functions present only in the metagenome; grey cells: functionspresent in both metagenome and metaproteome).

Symbiotic sustain within lichen-associated bacteriaM Grube et al

420

The ISME Journal

protein degradation of which 957 contigs werederived from bacteria. The cultivation-based functionalscreening underlines that Lobaria-associated bacteriaare key-players in the recycling of old lichen parts:33% of the isolates demonstrated protease activity,while 6.7 and 12.9% have shown chitinase andglucanase activity, respectively.

Other functions. In the KEGG analysis, thebiosynthesis of polyketides and terpenoids wassuggested in 2975 contigs, and the biosynthesis of othersecondary metabolites is suggested by a further 2012contigs. This set also includes known antibioticssuch as betalain, penicillin, streptomycin, butirosin,neomycin and novobiocin, as well as phenylpropa-noids, stilbenoids and alkaloids. A substantialnumber of contigs in the KEGG analysis was linkedwith pathways of environmental informationprocessing (12 053 contigs), and of these, membranetransport forms a fraction of 7629 contigs. Somebacteria also undergo CO2-fixation, of whichabout 757 bacterial contigs were assigned. Notably,polyol utilization is represented in 301 contigs(ribitol-, xylitol-, arabitol-, mannitol- and sorbitol-utilization).

Quorum sensing was rather poorly represented inour dataset with only 25 contigs. Metaproteomeanalysis supports the under-represented role ofquorum sensing within the Lobaria microbiome,no relevant sequences were identified. It was there-fore not surprising, that cultivation-dependentexperiments involving lichen-associated bacteriarevealed that o1% of tested isolates produceddetectable N-acyl homoserine–lactone-derivedquorum sensing molecules. In addition, phages,prophages and transposable elements were notparticularly common among the lichen-associatedmicrobiome (275 contigs).

Comparison with an algal partner. To confirmthat the selected functions noted above are trulydominant in the lichen-associated bacteria, wehave compared the SEED function analysis of theLobaria metagenome with that of a transcriptomeof D. reticulata, the green-algal photobiont ofL. pulmonaria. The transcriptome data wereobtained from the Joint Genome Institute repository(sequencing of three cultured partners of L. pulmo-naria, http://genome.jgi.doe.gov/Lobpulcupartners/Lobpulcupartners.info.html). There are a total of102 102 contigs in the assembly (about 27.71% of the368 424 contigs of the lichen metagenome). TheSEED analysis for the D. reticulata data wasperformed in exactly the same way as was done forthe lichen metagenome. Supplementary Table S2shows a comparison of the selected microbiomefunction assignments between these two datasets,whereas Supplementary Table S3 shows thesame kind of comparison of top-level SEEDfunctions. As Supplementary Table S2 clearlyshows, for every selected function discussed

until now, the lichen data has far more contigsassigned than the D. reticulata data has. Although agenomic/transcriptomic comparison with both fungaland algal partners is currently not feasible, somebacterial functions, which contribute greatly to theoverall symbiosis, have been confirmed.

Discussion

The analysis of our data has revealed metaboliccapacities and potential roles of the lichen-asso-ciated bacteria, especially in the areas of stabilityand survival of the overall symbiosis. Using ourmultiphasic approach, combining omics technolo-gies and physiological assays, we found diverse andpreviously unknown potential functions of themicrobiome, such as nutrient supply, resistanceagainst biotic and abiotic stress, support for thephotosynthesis and for the growth of the twoeukaryotic partners, as well as detoxification andthallus degradation abilities. Supportive roles ofassociated microbiomes are well-known fromhumans, animals and plants (Berg, 2009; Braginaet al., 2013; Cho and Blaser, 2012). With our presentresults, we have found new hints that a similarhelper effect can be present in lichen symbioses.This supports our concept of lichens as complexmicrobial ecosystem (Grube et al., 2009). Combiningmetagenomic analysis of a bacteria-enriched Lobariasample with a metaproteomic approach provided anadditional insight into the functional and structuraldiversity of bacterial inhabitants. The identifiedpotential functions of the lichen microbiome suggestinteractions with the algal as well as with the fungalpartner. Metagenomic and proteomic experimentshave provided evidence for the capability of pro-duction of vitamin B12 and other cofactors support-ing the beneficial algae–bacteria interaction. Manyalgae are auxotroph for vitamin B12, which is oftensynthesized by prokaryotes in symbiotic interac-tions (Croft et al., 2005). Other functions such asnutrient supply and resistance against biotic andabiotic stress factors indicate interactions with thefungal partner. Lichens are exposed to abiotic stressand a well-known target for parasitic fungi (Grubeet al., 2012). In addition, hormones such as auxinproduced by bacteria can support the growth ofalgae as well as fungi (Gutjahr, 2014). Althoughresults confirmed by both metagenomics and pro-teomics data corroborated functionalities, a moredetailed view is often limited by the availability ofannotated data and the lack of completelysequenced genomes, for comparison. Some differ-ences between the two approaches might also beexplained by variation of the actual metabolicactivity of the involved microorganisms at the timeof sample preparation. However, the present resultsagree with a previous metaproteomic approach of anindependent sample of the same lichen, coveringthe entire lichen holobiome. Schneider et al. (2011)

Symbiotic sustain within lichen-associated bacteriaM Grube et al

421

The ISME Journal

showed that algal proteins are involved in energyproduction and a diverse set of functions of thefungal proteins relates to the role of the mycobiontin energy consumption and symbiotic control.Similar findings resulted from the analysis of theeukaryotic metatranscriptome in another lichensymbiosis (Juntilla and Rudd, 2012; Juntilla et al.,2013). The present study extends the previousstudies in resolution of bacterial functionality.Focusing on the bacterial fraction, the comparisonof the new proteomic and metagenomic datasetsrevealed an overall similarity in the taxonomicrepresentation of bacterial organisms. In both data-sets, Proteobacteria are the most prominent phylum,with Alphaproteobacteria as the most prevalentclass. In contrast, cyanobacterial proteins weremuch more abundant as expected from the contigcounts; they might be more active than otherphylogenetic groups. As potential carbon and nitro-gen fixing organisms, they are known for a long timeas substantial part of lichen symbioses (Honeggeret al., 2013).

Bacterial communities on long-living lichenthallus surface remain largely constant over seasons(Grube et al., 2009), despite they are exposed tosubstantial periodicity of abiotic factors in thehabitat. Our data show that bacteria living on thesurfaces of lichens are well-adapted to abiotic stress,in particular osmotic and oxidative stress. Theseproperties match with the general property oflichens to tolerate periodic drought in their naturalhabitats. The periodic desiccation and rehydrationcycles lead to repeated oxidative bursts at thesurfaces of the fungal textures. Release of freeradicals under these circumstances has been demon-strated and was interpreted as pathogen defensemechanism of lichens (Minibayeva and Beckett,2001; Beckett et al., 2013). Thus, thallus-colonizingthalli bacteria without pronounced tolerance tooxidative stress and other selective conditionsbarely survive. By oxidative degradation of thesenon-adapted bacteria, a broad spectrum of addi-tional nutrients is accessible. This source might bemore important for oligotrophic lichens than forsoil-provisioned plants (Paunfoo-Lonhienne et al.,2010; White et al., 2012). Consequently we hypothe-size that periodic hydration acts as selective pres-sure for enrichment of specific and stress-tolerantbacterial communities, which can contribute tolongevity and persistence of lichens under extremeand changing ecological conditions.

As we found little evidence of quorum sensing,we hypothesize that bacterial colonization of thethallus is mostly regulated by the fungal partner inthe symbiotic community. It is well-established thatthe secondary metabolites of diverse lichen specieshave broad antibacterial properties (Boustie andGrube, 2005). The surprising abundance of bacteriaon the surfaces and between crystals of secondarymetabolites in lichens (for example, Lecanora poly-tropa; Grube et al., 2009) can only be explained

by differences in the susceptibility to antibioticcompounds, which may be considered anotherfactor of bacterial selection in lichens. Because wefound significant numbers of multidrug resistanceefflux pumps, the phylogenetically old lichensymbiosis could represent a natural reservoir ofbacterial resistance mechanisms. Moreover, some ofthe adapted bacteria are potentially involved in thedegradation of fungal secondary metabolites, asindicated by contigs of genes whose products mayprocess complex and cyclic carbohydrates. Thesegenes might also be interesting for biotechnologicalapproaches, aiming at the degradation of xenobiotics.The presence of genes for the metabolism of typicalbacterial antibiotics in our dataset suggests potentialcompetition among bacterial strains on the lichensurfaces or a defense against other strains enteringthe microbial surface community. The ecologicalsignificance of these functions is pending furtherexperimentation.

The morphological design of lichen structurescould have a profound effect on the organization ofthe symbiotic networking. Bacterial communitiesprimarily colonize the (hydrophilic) lichen surfaces,yet this pattern is strikingly different from theinternalized symbionts in lichens, such as the greenalgae (D. reticulata in L. pulmonaria), whichprimarily contribute to the provision of photo-synthetically produced carbohydrates. The green-algal strain is massively enriched within the fungalstructures, while cyanobacterial Nostoc strains areacquired from the surfaces of L. pulmonaria recur-rently during the life-time of the thallus to forminternal organs devoted to nitrogen fixation inlichens (Hyvarinen et al., 2002, Cornejo andScheidegger, 2013). Conversely, the external pre-sence of other bacteria in the lichen symbiosisrecalls the helper bacteria of mycorrhizal symbioses,which provide multiple functions to mutuallysupport and stabilize the root symbioses, includingexchange of carbohydrates and vitamin provision(Frey-Klett et al., 2007; Deveau et al., 2010). Thelong-living lichen thallus is formed by tightlypacked fungal hyphae, which are conglutinated bytheir cell walls. Lichen-adapted bacteria benefitfrom the persistent cell walls for nutrition, and inreturn provide multiple helper functions for thelongevity of lichen thalli (Figure 2b) to enhancefitness of the holobiome (symbiome). Although wehave found a lot of supportive facts for thissymbiosis model within our datasets, the evidencefor fulfilled contribution of bacterial communitiesto the lichen symbiosis can only be found byadditional experiments, for example, using isotope-labeled compounds and/or comparative physiologicalanalysis between lichens with and without thebacterial microbiota. Both are currently difficult toestablish due to slow metabolism and high diversityof bacteria. Despite these facts, we consider lichensas an interesting model for multi-biont symbioses,with different distributions of functions among

Symbiotic sustain within lichen-associated bacteriaM Grube et al

422

The ISME Journal

the symbionts. It appears that partners withnarrowly specified functions, such as carbohydrateor nitrogen fixation, are preferentially internalizedby fungal structures and massively enriched in thelichen thallus. Internalization of a partner may helpto provide a more uniform and stable environmentfor these partners. This symbiotic design hasevolved as a convergence in unrelated fungallineages (Grube and Hawksworth, 2007; Honegger,2012). It has been optimized not only for theassociation with carbon-providing algal photobionts(Kranner et al., 2005), but also for the enrichment ofbacterial supporters. Genome sequences of theeukaryotic partners, which are now becoming avail-able (for example, Wang et al., 2014), will furtherhelp to address the intricacies of one of the oldestknown symbiosis and its interactions with theirbacterial helpers.

Conflict of Interest

The authors declare no conflict of interest.

Acknowledgements

This work was supported by a joint grant of the Austrianand German Science Foundation (FWF, DFG) to GB, MGand KR (FWF-DACH Project I882).

References

Altschul SF, Gish W, Miller W, Myers EW, Lipman DJ.(1990). Basic local alignment search tool. J Mol Biol215: 403–410.

Bates ST, Cropsey GWG, Caporaso JG, Knight R, Fierer N.(2011). Bacterial communities associated withthe lichen symbiosis. Appl Environ Microbiol 77:1309–1314.

Beckett RP, Zavarzina AG, Liers C. (2013). Oxidoreduc-tases and cellulases in lichens: possible roles in lichenbiology and soil organic matter turnover. Fungal Biol117: 431–438.

Berg G. (2009). Plant-microbe interactions promoting plantgrowth and health: perspectives for controlled useof microorganisms in agriculture. Appl MicrobiolBiotechnol 84: 11–18.

Berg G, Roskot N, Steidle A, Eberl L, Zock A, Smalla K.(2002). Plant-dependent genotypic and phenotypicdiversity of antagonistic rhizobacteria isolatedfrom different Verticillium host plants. Appl EnvironMicrobiol 68: 3328–3338.

Boustie J, Grube M. (2005). Lichens—a promising sourceof bioactive secondary metabolites. Plant Gen Res 3:273–287.

Bragina A, Berg C, Muller H, Moser D, Berg G. (2013).Insights into functional bacterial diversity and itseffects on Alpine bog ecosystem functioning. Sci Rep3: 1995.

Cardinale M, Muller H, Berg G, de Castro Jr JV, Grube M.(2008). In situ analysis of the bacteria communityassociated with the reindeer lichen Cladonia

arbuscula reveals predominance of Alphaproteobac-teria. FEMS Microbiol Ecol 66: 63–71.

Cardinale M, Steinova J, Rabensteiner J, Berg G, Grube M.(2012a). Age, sun, and substrate: triggers of bacterialcommunities in lichens. Environ Microbiol Rep 4:23–28.

Cardinale M, Grube M, de Castro Jr JV, Muller H, Berg G.(2012b). Bacterial taxa associated with the lung lichenLobaria pulmonaria are differentially shaped bygeography and habitat. FEMS Microb Lett 329:111–115.

Cho I, Blaser MJ. (2012). The human microbiome: at theinterface of health and disease. Nat Rev Gen 13:260–270.

Cornejo C, Scheidegger C. (2013). New morphologicalaspects of cephalodium formation in the lichenLobaria pulmonaria (Lecanorales, Ascomycota).Lichenologist 45: 77–87.

Croft MT, Lawrence AD, Raux-Deery E, Warren MJ,Smith AG. (2005). Algae acquire vitamin B12 througha symbiotic relationship with bacteria. Nature 438:90–93.

Deveau A, Brule C, Palin B, Champmartin D, Rubini P,Garbaye J et al. (2010). Role of fungal trehalose andbacterial thiamine in the improved survival andgrowth of the ectomycorrhizal fungus Laccaria bicolorS238N and the helper bacterium Pseudomonas fluor-escens BBc6R8. Environ Microbiol Rep 4: 560–568.

Eddy SR. (2011). Accelerated profile HMM searches. PLoSComput Biol 7: e1002195.

Frank AB. (1877). Uber die biologischen Verhaltnisse desThallus einiger Krustenflechten. Beitr Biol Pflanzen 2:123–200.

Frey-Klett P, Garbaye J, Tarkka M. (2007). The mycorrhizahelper bacteria revisited. New Phytol 176: 22–36.

Grube M, Hawksworth DL. (2007). Trouble with lichen:the re-evaluation and re-interpretation of thallus formand fruit body types in the molecular era. Mycol Res111: 1–1132.

Grube M, Cardinale M, Vieira de Castro J, Muller H,Berg G. (2009). Species-specific structural andfunctional diversity of bacterial communities in lichensymbioses. ISME J 3: 1105–1115.

Grube M, Koberl M, Lackner S, Berg C, Berg G. (2012).Host-parasite interaction and microbiome response:effects of fungal infections on the bacterial communityof the Alpine lichen Solorina crocea. FEMS MicrobiolEcol 82: 472–481.

Gutjahr C. (2014). Phytohormone signaling in arbuscularmycorhiza development. Curr Opin Plant Biol 21:26–34.

Hai B, Diallo NH, Sall S, Haesler F, Schauss K, Bonzi Met al. (2009). Quantification of key genes steeringthe microbial nitrogen cycle in the rhizosphere ofsorghum cultivars in tropical agroecosystems. ApplEnviron Microbiol 75: 4993–5000.

Haft DH, Selengut JD, Richter RA, Harkins D, Basu MK,Beck E. (2013). TIGRFAMs and genome properties in2013. Nucleic Acids Res 41(Database issue): D387–D395.

Hodkinson BP, Gottel NR, Schadt CW, Lutzoni F. (2011).Photoautotrophic symbiont and geography aremajor factors affecting highly structured and diversebacterial communities in the lichen microbiome.Environ Microbiol 14: 147–161.

Honegger R. (2012). The symbiotic phenotype of lichen-forming Ascomycetes and their endo- and epibionts.

Symbiotic sustain within lichen-associated bacteriaM Grube et al

423

The ISME Journal

In: Hock B (ed.) The Mycota IX. Springer: Berlin, pp287–339.

Honegger R, Edwards D, Axe L. (2013). The earliestrecords of internally stratified cyanobacterial andalgal lichens from the Lower Devonian of the WelshBorderland. New Phytol 197: 264–275.

Huson DH, Mitra S, Weber N, Ruscheweyh H,Schuster SC. (2011). Integrative analysis of environ-mental sequences using MEGAN4. Genome Res 21:1552–1560.

Hyvarinen M, Hardling R, Tuomi J. (2002). Cyanobacteriallichen symbiosis: the fungal partner as an optimalharvester. Oikos 98: 498–502.

Juntilla S, Rudd S. (2012). Characterization of atranscriptome from a non-model organism, Cladoniarangiferina, the grey reindeer lichen, using high-throughput next generation sequencing and ESTsequence data. BMC Genomics 13: 575.

Junttila S, Laiho A, Gyenesei A, Rudd S. (2013).Whole transcriptome characterization of the effectsof dehydration and rehydration on Cladonia rangifer-ina, the grey reindeer lichen. BMC Genomics 14: 870.

Kranner I, Cram WJ, Zorn M, Wornik S, Yoshimura I,Stabentheiner E, Pfeifhofer HW. (2005). Antioxidantsand photoprotection in a lichen as compared with itsisolated symbiotic partners. PNAS 102: 3142–3146.

Koskinen VR, Emery PA, Creasy DM, Cottrell JS. (2011).Hierarchical clustering of shotgun proteomics data.Mol Cell Proteomics 10: M110.003822.

Kuester E, Williams ST. (1964). Selection of media forisolation of Streptomycetes. Nature 202: 928–929.

Laemmli UK. (1970). Cleavage of structural proteinsduring the assembly of the head of bacteriophage T4.Nature 227: 680–685.

Liba CM, Ferrara FIS, Mangio GP, Fantinatti-Garboggini F,Albuquerque RC, Pavan C et al. (2006). Nitrogen-fixingchemo–organotrophic bacteria isolated fromcyanobacteria-deprived lichens and their ability tosolubilize phosphate and to release amino acids andphytohormones. J Appl Microbiol 101: 1076–1086.

Margulis L, Fester R. (1991). Symbiosis as a source ofevolutionary innovation: speciation and morphogenesis.MIT Press: Cambridge, pp 454.

Martin BD, Schwab E. (2013). Current usage of symbiosisand associated terminology. Int J Biol 5: 32–45.

McClean KH, Winson MK, Fish L, Taylor A, Chhabra SR,Camara M et al. (1997). Quorum sensing andChromobacterium violaceum: exploitation of viola-cein production and inhibition for the detectionof N-acylhomoserine lactones. Microbiology 143:3703–3711.

Meyer F, Paarmann D, D’Souza M, Olson R, Glass EM,Kubal M et al. (2008). The metagenomics RASTserver - a public resource for the automatic phyloge-netic and functional analysis of metagenomes. BMCBioinformatics 9: 386.

Minibayeva F, Beckett RP. (2001). High rates of extra-cellular superoxide production in bryophytes andlichens, and an oxidative burst in response torehydration following desiccation. New Phytol 152:333–343.

Mushegian AA, Peterson CN, Baker CCM, Pringle A.(2011). Bacterial diversity across individual lichens.Appl Environ Microbiol 77: 4249–4252.

Nash III TH. (2008). Lichen Biology, 2nd edn. CambridgeUniversity Press: Cambridge, UK.

Nautiyal CS. (1999). An efficient microbiological growthmedium for screening phosphate solubilizing micro-organisms. FEMS Microbiol Lett 170: 265–270.

Neuhoff V, Arold N, Taube D, Ehrhardt W. (1988).Improved staining of proteins in polyacrylamide gelsincluding isoelectric focusing gels with clear back-ground at nanogram sensitivity using CoomassieBrilliant Blue G-250 and R-250. Electrophoresis 9:255–262.

Opelt K, Chobot V, Hadacek F, Schonmann S, Eberl L,Berg G. (2007). Investigations of the structure andfunction of bacterial communities associated withSphagnum mosses. Environ Microbiol 9: 2795–2809.

Øvstedal D, Lewis-Smith R. (2001). Lichens of Antarcticaand South Georgia, A guide to their identification andecology. Cambridge University Press: Cambridge, UK.

Paungfoo-Lonhienne C, Rentsch D, Robatzek S, Webb RI,Sagulenko E, Nasholm T et al. (2010). Turning thetable: plants consume microbes as a source ofnutrients. PLoS One 5: e11915.

Scheidegger C, Werth S. (2009). Conservation strategies forlichens: insights from population biology. Fungal BiolRev 23: 55–66.

Schneider T, Schmid E, de Castro Jr JV, Cardinale M,Eberl L, Grube M et al. (2011). Structure and functionof the symbiosis partners of the lung lichen (Lobariapulmonaria L. Hoffm.) analyzed by metaproteomics.Proteomics 11: 2752–2756.

Shirling EB, Gottlieb D. (1966). Methods for characteriza-tion of Streptomyces species. Int J Syst Bacteriol 16:313–340.

Steidle A, Sigl K, Schuhegger R, Ihring A, Schmid M,Gantner S et al. (2001). Visualization of N-acylhomo-serine lactone-mediated cell-cell communicationbetween bacteria colonizing the tomato rhizosphere.Appl Environ Microbiol 67: 5761–5770.

Wang W, Vignani R, Scali M, Cresti M. (2006). A universaland rapid protocol for protein extraction fromrecalcitrant plant tissues for proteomic analysis.Electrophoresis 27: 2782–2786.

Wang YY, Liu B, Zhang XY, Zhou QM, Zhang T, Li H et al.(2014). Genome characteristics reveal the impact oflichenization on lichen-forming fungus Endocarponpusillum Hedwig (Verrucariales, Ascomycota). BMCGenomics 15: 34.

White Jr JF, Crawford H, Torres MS, Mattera R, Irizarry I,Bergen M. (2012). A proposed mechanism for nitrogenacquisition by grass seedlings through oxidation ofsymbiotic bacteria. Symbiosis 57: 161–171.

Zerbino DR, Velvet BE. (2008). Algorithms for de novoshort read assembly using de Bruijn graphs. GenomeRes 18: 821–829.

Zybailov B, Mosley AL, Sardiu ME, Coleman MK,Florens L, Washburn MP. (2006). Statisticalanalysis of membrane proteome expression changesin Saccharomyces cerevisiae. J Proteome Res 5:2339–2347.

Supplementary Information accompanies this paper on The ISME Journal website (http://www.nature.com/ismej)

Symbiotic sustain within lichen-associated bacteriaM Grube et al

424

The ISME Journal